An electronic spreadsheet allows a user to enter a combination of data and formulaic relationships between the data into a matrix. Each cell of the matrix holds either a single data item or a formula describing the computed relationship to one or more other cells. The power of the electronic spreadsheet becomes readily apparent when one realizes that any given computed cell can serve as input data to another computed cell. This allows for financial data analysis models of arbitrary complexity to be constructed with any number of possible benefits to the user.
The flexibility and unstructured nature of a typical conventional electronic spreadsheet application may cause data modeling relationships to become exceedingly difficult to understand, even by the original author of the spreadsheet. Additionally, changes to one part of the data model can lead to unexpected breakdowns elsewhere in the spreadsheet. Enacting counter-measures against such breakdowns is time-consuming, tedious and error-prone.
Additionally, for all of the computational capabilities of a typical spreadsheet application, the resulting analysis that can be performed is typically limited to changes in one or two independent variables at a time and then visually observing a few dependent variables downstream. In a complex data model, such as those found in a conventional electronic financial model, there may be many dependent and independent variables. Thus, it may be significantly challenging to be able to modify all of these variables and see the resulting changes in and from numerous potential inputs. As a result, there may be a reduction in understanding of the model itself. Highly-derived values and complex high-order relationships between various aspects of the model become extremely difficult to discern.
Conventional electronic spreadsheets used in financial modeling which include a temporal component may be referred to as electronic planning systems. Conventional electronic planning systems may enable users to build sophisticated financial models with complex formulaic relationships between various aspects of the model. “Sensitivity analysis” as used herein describes the study of one or more calculation results deriving from one or more changes to an input variable. Conventional electronic planning systems typically express the terms of the inputs to the analysis, the formulae, and outputs from the analysis in short cryptic abbreviated terms. This makes sensitivity analysis cumbersome and confusing because the short cryptic abbreviated terms may not clearly express their meaning in natural language terms that can be easily understood. As a result, there may be a reduction in understanding of the financial model itself.
To overcome this, users often construct elaborate copies of input and output data to transform one or both data sets into terms more conducive to the desired analysis. Such a technique is tedious, error prone, can hide resulting calculation flaws, and cannot provide the needed sensitivity analysis. The workflow involved makes ad hoc analysis and data exploration next to impossible.